Days to Liquidate
Volatility says how much a position can move. Days to liquidate says how long it would take to leave it without becoming a large part of the market for that security. For a fund that may have to meet redemptions, the second question can matter more than the first.
Overview
Every stock run reports a liquidity block for each method. It uses the traded volume that comes with the price history, so it needs no separate data source. A mutual fund run has no traded volume and reports no block, and neither does a run whose volume history is unusable. The block is then absent, which means unknown. It is never a block of zeros, which would mean something different.
A run holds weights, not money. The block is therefore sized on a stated notional of 10,000,000 in the run's currency, which is one crore rupees on an Indian run. Exit time scales linearly with position size, so for a portfolio of a different size you multiply every figure by the ratio of the two values.
Mathematical Formulation
Average daily value
The normal daily traded value of security is the median of close price times volume over the last 63 sessions of the sample, counting only sessions with a positive value:
The median, not the mean, because one block trade can be many times a normal day's volume, and a mean would let a single print make an illiquid security look tradeable for a whole quarter. Value traded, not share count, because a portfolio is sized in money and a share count is not comparable across prices.
Days to liquidate
With portfolio value , weight and participation rate , the share of a normal day one seller can take without being the reason the price moves:
A participation rate of 20% is a common trading-desk limit. It is an assumption, so the block reports it beside every figure derived from it.
The portfolio figure
A portfolio exits as slowly as its slowest holding, so the headline is the maximum, not the average:
Here is the set of holdings with a usable ADV. The weighted mean is context. Reporting only the average is how an illiquid tail is hidden.
Interpretation
| Band | Worst holding | Reading |
|---|---|---|
comfortable | Every position exits within one session at the stated rate. | |
moderate | At least one position needs several sessions. | |
concentrated | At least one position could decide whether an orderly exit is possible. | |
unknown | not finite | No figure could be computed. |
The band is a label. The number is the result: read the days, and rescale them to your own portfolio value.
Fields
| Field | Meaning |
|---|---|
participation_rate | The assumed share of a normal day's traded value, 0.20. |
portfolio_value | The notional the figures are sized on: 10,000,000 in the run's currency. |
adv_window_days | The trailing window of the ADV median, 63 sessions. |
adv_statistic | Always median. |
average_daily_value | The ADV of each covered holding, in money per day. |
days_to_liquidate | The exit time of each covered holding, in trading days. |
worst_holding | The holding with the longest exit time. |
worst_days_to_liquidate | That exit time: the headline figure. |
weighted_days_to_liquidate | The weight-averaged exit time over the covered holdings. |
uncovered_holdings | Holdings with no usable volume, so no figure. |
band | comfortable, moderate, concentrated or unknown. |
Limitations
- A measure of the market, not of an execution. It says nothing about market impact, the bid-ask spread or the price a trade would get. No impact model is applied, because an uncalibrated cost estimate is worse than none.
- Recent, not stressed. The window is the last quarter of the sample. Traded value falls in a sell-off, which is when an exit is most likely to be needed, so the figure understates exit time in stress.
- One venue. The volume is that of the listing the run priced. A security traded on both NSE and BSE has more liquidity than one listing shows.
- Uncovered holdings are not liquid. A holding without usable volume is listed in
uncovered_holdings. Absence of a figure is not a figure of zero.
The Fund Manager Mode page states the same boundary: the liquidity measures report how long an exit would take at a stated participation rate, and they do not price the trade.
References
- Amihud, Y. (2002). "Illiquidity and Stock Returns: Cross-Section and Time-Series Effects." Journal of Financial Markets, 5(1), 31-56. doi:10.1016/S1386-4181(01)00024-6.
- Almgren, R., & Chriss, N. (2001). "Optimal Execution of Portfolio Transactions." Journal of Risk, 3(2), 5-39. doi:10.21314/JOR.2001.041.
- Basel Committee on Banking Supervision (2019). Minimum Capital Requirements for Market Risk (the liquidity horizons of the revised market risk framework). Bank for International Settlements. bis.org/bcbs/publ/d457.htm.
Not investment advice. Past performance is not indicative of future results.